In view of the low contrast and many noise points of the mesh fabric image, the segmentation results have the problems of mesh connection and incompleteness. A segmentation algorithm based on the minimum gray value of the region is proposed to improve the segmentation accuracy of the mesh. First, the image is processed with Gaussian pyramid scaling and histogram equalization algorithm to enhance the texture contour and contrast of light and dark of the image. Then, a segmentation algorithm based on the minimum gray value of the area is used to solve the problem that the mesh cannot be segmented correctly only by the gray value. Finally, a multi-image fusion algorithm is used to solve the problem of difficult threshold selection in the segmentation algorithm based on local gray scale minima. A variety of mesh fabric images with different illumination levels are selected for experiments. The experimental results show that the proposed algorithm has a good segmentation effect, which can effectively solve the problems of mesh adhesion and incompleteness in the segmentation results. The segmentation error rate of the mesh fabric is 0.24%.